ATLAS Data API Mechanism

a generic web interface for the retrieval and display of the ATLAS TDAQ information

The ADAM project goal is to reduce the time an expert needs to investigate a problem through the correlation of various statistics. A plug and play mechanism allows any provider to connect to the system, advertise capabilities, and expose its data using a flexible communication protocol named the ADAM interface.

The Vision1. Design an abstraction layer to deal with

data retrieval from all the monitoring applications (the ADAM interface);

2. Build a correlation and aggregation engine to mix data from any provider according to user needs;

3. Develop a web-based visualization interface to display raw or aggregated data in a meaningful way. The interface was inspired from the Google Visualization project.

MRSAnalytics

ADAM

INTERFACE

(L0)

NET-IS

……

COOL

The ADAM interface• A flexible and highly customizable REST-full interface to transfer formatted data between a server and a client. Data communication is done via HTTP;

• A client has to use a SQL-like query language to ask for provider data;

• Support for various data delivery packaging: XML, JSON, CSV and HTML.

ADAM INTERFACE (L1)

The visualization portal

The Data• The ATLAS monitoring systems collects various type of information (e.g. time-series statistics, event-triggered messaged etc) and use many storage technologies for the collected data (e.g. relational databases, RRD files, custom formats etc);

• Some systems have custom interfaces to expose their data to other applications while others have none;

• While the storage technology is an internal aspect of a monitoring system (provider), all the data types recorded by the ATLAS TDAQ monitoring tools are supported by the ADAM interface.

Harwood Adam ¹, Lehmann Miotto Giovanna ², Magnoni Luca ², Savu Dan ², Vandelli Wainer ²

¹ Adam HOME INSTITUTE, ² CERN

Query Example:http://server/adam?query=SELECT provider:table:column FROM Analytics &op:out=html

DATA PROCESSING ENGINE

CLI client

• User customizable web-based interface;• Single source or multiple source support, through the data processing engine.• Easy-to-interpret charts;• Real-time, interactive graphs;•Adjustable and configurable chart displays;•Hierarchical plot pages navigation supporting user defined branches.

The data processing engine uses various techniques (e.g. correlation, aggregation, filtering, scaling, interpolation, pattern recognition etc) to process the data ,fed from various data providers, in an easy-to-display format to be sent to the user interface.

● Implemented as a java servlet;● Plug’n play support for providers;● Database backend;● The output is compliant with the ADAM interface specifications.

This research project has been supported by a Marie Curie Initial Training Network Fellowship of the European Community’s & Seventh Framework Programme under contract number (PITN-GA-2008-211801-ACEOLE) and by the European Community's Seventh Framework Programme under the Grant Agreement no 212114 (SLHC-PP)

Webbrowser

HTTP HTTP

ALL IN ONE VIEW The ATLAS experiment is the largest particle physics experiment ever built. Many systems have been deployed to monitor the detector and the dataflow while running. Most of these systems have their own set of log messages and statistics provided through various custom user interfaces.